Exercise 1: American Airlines Employees

# loading libraries

library(tibble)
library(dplyr)
library(ggplot2)
library(ggpubr)
library(scales)
library(lubridate)

# plot size

std_width  <- 14
std_height <- 9
options(repr.plot.width = std_width, repr.plot.height = std_height)

# colors for plot

color_palette <- c( "#58508d",
                    "#bc5090",
                    "#ff6361",
                    "#ffa600" )

1) Read and import the data

# import data 

american <- read.table( "american_airline_empl.txt", header = T, sep = "\t" )
delta    <- read.table( "delta_airline_empl.txt",    header = T, sep = "\t" )
fedex    <- read.table( "federal_express_empl.txt",  header = T, sep = "\t" )
united   <- read.table( "united_airline_empl.txt",   header = T, sep = "\t" )

2) Create a common tibble

# add airline column to each dataset

american <- american %>% add_column(Company = "American Airlines")
delta <- delta %>% add_column(Company = "Delta Airlines")
fedex <- fedex %>% add_column(Company = "Federal Express Airlines")
united <- united %>% add_column(Company = "United Airlines")

# create unique tibble

data <- rbind(american, delta, fedex, united)
data <- as_tibble(data)
head(data)

A tibble: 6 × 6

Month	Year	Full.time	Part.time	Grand.Total	Company
<int>	<int>	<chr>	<chr>	<chr>	<chr>
1	1990	68,137	9,039	77,176	American Airlines
2	1990	68,725	9,273	77,998	American Airlines
3	1990	69,509	9,376	78,885	American Airlines
4	1990	69,713	9,326	79,039	American Airlines
5	1990	70,376	9,309	79,685	American Airlines
6	1990	71,258	9,369	80,627	American Airlines

# adjusting tibble (removing commas)

data <- data %>% transmute( 
                Month = Month,
                Year = Year,
                FT = as.numeric(sub(",", "", Full.time, fixed = TRUE)),
                PT = as.numeric(sub(",", "", Part.time, fixed = TRUE)),
                Tot = as.numeric(sub(",", "", Grand.Total, fixed = TRUE)),
                Company = Company )

3) Plot of the employees as a function of time for all companies

# calculating number of employees annual mean for each company 

dataM <- data %>% group_by(Year, Company)
# filtering out year 2023 because in this case we don't need it
dataM <- filter(dataM, Year != 2023) 
dataM <- dataM %>% summarise(FT = mean(FT), PT = mean(PT), Tot = mean(Tot))

`summarise()` has grouped output by 'Year'. You can override using the
`.groups` argument.

head(dataM)

A grouped_df: 6 × 5

Year	Company	FT	PT	Tot
<int>	<chr>	<dbl>	<dbl>	<dbl>
1990	American Airlines	71252.75	9463.417	80716.17
1990	Delta Airlines	57445.75	4301.500	61747.25
1990	Federal Express Airlines	64834.25	23565.250	88399.50
1990	United Airlines	67644.58	5690.667	73335.25
1991	American Airlines	77587.92	11084.583	88672.50
1991	Delta Airlines	62714.42	4951.000	67665.42

# full-time plot

gFT <- ggplot( dataM, aes(x = Year, y = FT, color=Company) ) +
       geom_point( size = 3 ) +
       geom_line( linewidth = 1 )

gFT <- gFT + 
       theme_bw() +
       theme( legend.position = "right",
        plot.title = element_text(size = 26, hjust = 0.5),
        axis.title = element_text(size = 20),
        axis.text  = element_text(size = 18),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                          linewidth = 0.3,
                                          linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Year", 
              y = "Full-Time Employees", 
              title = "Full-Time Employees Annual Averages" )

gFT <- gFT +
        scale_x_continuous(breaks = seq(1990, 2024, by = 5)) + 
        scale_y_continuous(n.breaks = 7)

# part-time plot

gPT <- ggplot( dataM, aes(x = Year, y = PT, color=Company) ) +
       geom_point( size = 3 ) +
       geom_line( linewidth = 1 )

gPT <- gPT + 
       theme_bw() +
       theme( legend.position = "right",
        plot.title = element_text(size = 26, hjust = 0.5),
        axis.title = element_text(size = 20),
        axis.text  = element_text(size = 18),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           size = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                          size = 0.3,
                                          linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Year", 
              y = "Part-Time Employees", 
              title = "Part-Time Employees Annual Averages" )

gPT <- gPT +
        scale_x_continuous( breaks = seq(1990, 2024, by = 5) ) + 
        scale_y_continuous( breaks = seq(0, 70000, by = 10000) )

gFT
gPT

# day by day

dataT <- data %>% group_by(Company)
dataT <- dataT %>% transmute(
                    # days sequence
                    Time = seq( as.Date("1990/01/01"), as.Date("2023/01/01"), "month" ), 
                    FT = FT, 
                    PT = PT,
                    Tot = Tot,
                    Company = Company )

# plot FT

g3a <- ggplot(dataT, aes(x = Time, y = FT, color = Company)) +
       geom_line(linewidth = 1.5 ) 

g3a <- g3a + 
       theme_bw() +
       theme( legend.position = "right",
        plot.title = element_text(size = 26, hjust = 0.5),
        axis.title = element_text(size = 20),
        axis.text  = element_text(size = 18),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           size = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                          size = 0.3,
                                          linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Time", 
              y = "Full-Time Employees", 
              title = "Full-Time Employees in Years 1990 - 2023" )

g3a <- g3a +
        scale_x_date(date_breaks = "3 year", date_labels = "%Y") +
        scale_y_continuous( n.breaks = 9 )

# plot PT

g3b <- ggplot(dataT, aes(x = Time, y = PT, color = Company)) +
       geom_line(linewidth = 1.5 ) 

g3b <- g3b + 
       theme_bw() +
       theme( legend.position = "right",
        plot.title = element_text(size = 26, hjust = 0.5),
        axis.title = element_text(size = 20),
        axis.text  = element_text(size = 18),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           size = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                          size = 0.3,
                                          linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Time", 
              y = "Part-Time Employees", 
              title = "Part-Time Employees in Years 1990 - 2023" )

g3b <- g3b +
        scale_x_date(date_breaks = "3 year", date_labels = "%Y") +
        scale_y_continuous( n.breaks = 9 )

g3a
g3b

4) Minimum and maximum number of employees for each company

# minimum number of employees

filter(dataT, Tot <= min(Tot))

A grouped_df: 4 × 5

Time	FT	PT	Tot	Company
<date>	<dbl>	<dbl>	<dbl>	<chr>
2013-09-01	55462	6828	62290	American Airlines
2006-11-01	41948	4462	46410	Delta Airlines
1990-01-01	61305	23580	84885	Federal Express Airlines
2011-06-01	40522	5259	45781	United Airlines

# maximum number of employees

filter(dataT, Tot >= max(Tot))

A grouped_df: 4 × 5

Time	FT	PT	Tot	Company
<date>	<dbl>	<dbl>	<dbl>	<chr>
2018-06-01	96543	12628	109171	American Airlines
2023-01-01	94236	439	94675	Delta Airlines
2021-03-01	204406	65977	270383	Federal Express Airlines
2001-03-01	91041	11005	102046	United Airlines

# plot to visualize min and max number of employees for each company
# plot of total number of employees trend over years for each company

gET <- ggplot( dataT, aes(x = Time, y = Tot, color = Company) ) +
       geom_line( linewidth = 1.3 )

gET <- gET + 
       theme_bw() +
       theme( legend.position = "right",
        plot.title = element_text(size = 26, hjust = 0.5),
        axis.title = element_text(size = 20),
        axis.text  = element_text(size = 18),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           size = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                          size = 0.3,
                                          linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Time", 
              y = "Total Employees", 
              title = "Total Employees in Years 1990 - 2023" )

gET <- gET +
        scale_x_date(date_breaks = "3 year", date_labels = "%Y") +
        scale_y_continuous( n.breaks = 9 )

gET

5) Plot of the fraction of PT worker over TOT employees as a function of time

# calculating PT/TOT fraction

dataF <- dataT %>% transmute( Time = Time,
                              PTF = PT / Tot,
                              Company = Company)

# plot

gPTF <- ggplot( dataF, aes(x = Time, y = PTF, color = Company) ) +
       geom_line( linewidth = 1.3 )

gPTF <- gPTF + 
       theme_bw() +
       theme( legend.position = "right",
        plot.title = element_text(size = 26, hjust = 0.5),
        axis.title = element_text(size = 20),
        axis.text  = element_text(size = 18),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           size = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                          size = 0.3,
                                          linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Time", 
              y = "% Part-Time Employees", 
              title = "Part-Time employees on Total Employees over Time" )

gPTF <- gPTF +
        scale_x_date(date_breaks = "5 years", date_labels = "%Y") +
        scale_y_continuous(labels = scales::percent)

gPTF

6) Influence of pandemic: trend in years 2019 - 2023

# analyzing years 2019-2023
dataP <- filter(dataT, Time >= "2019/01/01")

# plot

gTP <- ggplot( dataP, aes(x = Time, y = Tot, color = Company) ) +
       geom_point( size = 3 ) +
       geom_line( linewidth = 1 )

gTP <- gTP + 
       theme_bw() +
       theme( legend.position = "right",
        plot.title = element_text(size = 26, hjust = 0.5),
        axis.title = element_text(size = 20),
        axis.text  = element_text(size = 16),
        axis.text.x=element_text(angle=30, hjust=1),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           size = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                          size = 0.3,
                                          linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Time", 
              y = "Total Employees", 
              title = "Total Employees in Years 2019 - 2023" )

gTP <- gTP +
        scale_x_date(date_breaks = "6 month", date_labels = "%Y %b") +
        scale_y_continuous( n.breaks = 8 )

gTP

Apart from FedEx Airlines, a similar trend during years 2019 - 2023 is evident for the other three.

Globally, with the outbreak of the pandemic, from March 2020 the number of employees begins to decline to a relative minimum and then steadily rise again, returning, in January 2023, to about pre-pandemic levels. Specifically, Delta Airlines is the first company to reach the relative minimum, nearly halving the number of employees. In contrast, American Airlines and United Airlines follow a very similar trend, reaching a relative minimum of total employees in October 2020 and then slowly increasing. Moreover, the relative lows reached during the pandemic are not as low as the lows reached due to the effects of the 2007-2008 financial crisis.

These trends show that the pandemic has affected the number of total employees for the airlines analyzed (apart from FedEx). In fact, the number of total employees declined following the outbreak of the pandemic, reaching a low more or less in a short period of time, and then rising steadily until today.

As for FedEx Airlines, on the other hand, the trend is completely different, overall increasing since 1990. Moreover, for this company, the number of total employees increases rapidly from the outbreak of pandemic to the end of 2020.

---

Exercise 2: Data Frames and Tibble

• nycflights13 package

library(nycflights13)

1.1) Total number of flights departed from each of the three NYC airports as a function of time (one entry for each of the 365 days of the year)

head(flights)

A tibble: 6 × 19

year	month	day	dep_time	sched_dep_time	dep_delay	arr_time	sched_arr_time	arr_delay	carrier	flight	tailnum	origin	dest	air_time	distance	hour	minute	time_hour
<int>	<int>	<int>	<int>	<int>	<dbl>	<int>	<int>	<dbl>	<chr>	<int>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dttm>
2013	1	1	517	515	2	830	819	11	UA	1545	N14228	EWR	IAH	227	1400	5	15	2013-01-01 05:00:00
2013	1	1	533	529	4	850	830	20	UA	1714	N24211	LGA	IAH	227	1416	5	29	2013-01-01 05:00:00
2013	1	1	542	540	2	923	850	33	AA	1141	N619AA	JFK	MIA	160	1089	5	40	2013-01-01 05:00:00
2013	1	1	544	545	-1	1004	1022	-18	B6	725	N804JB	JFK	BQN	183	1576	5	45	2013-01-01 05:00:00
2013	1	1	554	600	-6	812	837	-25	DL	461	N668DN	LGA	ATL	116	762	6	0	2013-01-01 06:00:00
2013	1	1	554	558	-4	740	728	12	UA	1696	N39463	EWR	ORD	150	719	5	58	2013-01-01 05:00:00

fNYC <- flights %>% group_by(year, month, day, origin)
fNYC <- fNYC %>% summarise(tot = length(flight))
fNYC <- fNYC %>% transmute( date = paste(year, month, day, sep = "-"),
                            origin = origin,
                            tot = tot )
fNYC <- fNYC %>% transmute( date = as.Date(date),
                            Airport = origin,
                            tot = tot )
head(fNYC)

`summarise()` has grouped output by 'year', 'month', 'day'. You can override
using the `.groups` argument.

A grouped_df: 6 × 6

year	month	day	date	Airport	tot
<int>	<int>	<int>	<date>	<chr>	<int>
2013	1	1	2013-01-01	EWR	305
2013	1	1	2013-01-01	JFK	297
2013	1	1	2013-01-01	LGA	240
2013	1	2	2013-01-02	EWR	350
2013	1	2	2013-01-02	JFK	321
2013	1	2	2013-01-02	LGA	272

# plot

g <- ggplot( fNYC, aes( x = date, y = tot, color = Airport) ) + 
    geom_line( linewidth = 0.7) + 
    geom_point( size = 1)

g <- g +
    theme_bw() + 
    theme( legend.position = "right",
        plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Year", 
              y = "Total Number of Flights", 
              title = "Total Number of Flights from each of the three NYC airports during 2013" )

g <- g +
        scale_x_date(date_breaks = "1 month", date_labels = "%b") +
        scale_y_continuous( n.breaks = 10 )

g

1.2.1) Average number of flights computed over the first five working days of each week as a function of the week number of the year

fNYC <- fNYC %>% transmute( week = isoweek(ymd(date)),
                            date = date,
                            Airport = Airport,
                            tot = tot )
fNYC <- fNYC %>% transmute( week = week,
                            dow = wday(date, week_start = 1),
                            date = date,
                            Airport = Airport,
                            tot = tot )
head(fNYC)

A grouped_df: 6 × 8

year	month	day	week	dow	date	Airport	tot
<int>	<int>	<int>	<dbl>	<dbl>	<date>	<chr>	<int>
2013	1	1	1	2	2013-01-01	EWR	305
2013	1	1	1	2	2013-01-01	JFK	297
2013	1	1	1	2	2013-01-01	LGA	240
2013	1	2	1	3	2013-01-02	EWR	350
2013	1	2	1	3	2013-01-02	JFK	321
2013	1	2	1	3	2013-01-02	LGA	272

# average over first 5 working days

fNYCwd <- filter(fNYC, dow <=5)
fNYCwd <- fNYCwd %>% group_by(week, Airport)
fNYCwd <- fNYCwd %>% transmute( week = week,
                            dow = dow,
                            date = date,
                            Airport = Airport,
                            tot = mean(tot) )
head(fNYCwd)

A grouped_df: 6 × 5

week	dow	date	Airport	tot
<dbl>	<dbl>	<date>	<chr>	<dbl>
1	2	2013-01-01	EWR	325.6667
1	2	2013-01-01	JFK	308.1667
1	2	2013-01-01	LGA	259.1667
1	3	2013-01-02	EWR	325.6667
1	3	2013-01-02	JFK	308.1667
1	3	2013-01-02	LGA	259.1667

# plot 

gwd <- ggplot(fNYCwd, aes( x = week, y = tot, color = Airport) ) + 
        geom_line( linewidth = 1 ) + 
        geom_point( size = 1.6 )

gwd <- gwd +
    theme_bw() + 
    theme( legend.position = "right",
        plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Week", 
              y = "Average over Working Days", 
              title = "Average Number of Flights over Working Days from NYC Airports in 2013" )

gwd <- gwd +
        scale_x_continuous( n.breaks = 10 ) +
        scale_y_continuous( n.breaks = 10 )

gwd

# average on we

fNYCwe <- filter(fNYC, dow > 5)
fNYCwe <- fNYCwe %>% group_by(week, Airport)
fNYCwe <- fNYCwe %>% transmute( week = week,
                            dow = dow,
                            date = date,
                            Airport = Airport,
                            tot = mean(tot) )
head(fNYCwe)

A grouped_df: 6 × 5

week	dow	date	Airport	tot
<dbl>	<dbl>	<date>	<chr>	<dbl>
1	6	2013-01-05	EWR	269.5
1	6	2013-01-05	JFK	304.5
1	6	2013-01-05	LGA	202.0
1	7	2013-01-06	EWR	269.5
1	7	2013-01-06	JFK	304.5
1	7	2013-01-06	LGA	202.0

# plot 

gwe <- ggplot(fNYCwe, aes( x = week, y = tot, color = Airport) ) + 
        geom_line( linewidth = 1 ) + 
        geom_point( size = 1.6 )

gwe <- gwe +
    theme_bw() + 
    theme( legend.position = "right",
        plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Week", 
              y = "Average over Weekend", 
              title = "Average Number of Flights over Weekend from NYC Airports in 2013" )

gwe <- gwe +
        scale_x_continuous( n.breaks = 10 ) +
        scale_y_continuous( n.breaks = 10 )

gwe

2) Departure delay

dd <- flights %>% group_by(year, month, day, origin)
dd <- dd %>% transmute( date = paste(year, month, day, sep = "-"),
                        dep_delay = dep_delay,
                        flight = flight )
dd <- dd %>% transmute( date = as.Date(date),
                        depDelay = dep_delay,
                        flight = flight,
                        Airport = origin )

dd <- dd %>% ungroup()
#  filtering out NA in depDelay
dd <- dd %>% filter(!is.na(depDelay)) 

head(dd)

A tibble: 6 × 8

year	month	day	origin	date	depDelay	flight	Airport
<int>	<int>	<int>	<chr>	<date>	<dbl>	<int>	<chr>
2013	1	1	EWR	2013-01-01	2	1545	EWR
2013	1	1	LGA	2013-01-01	4	1714	LGA
2013	1	1	JFK	2013-01-01	2	1141	JFK
2013	1	1	JFK	2013-01-01	-1	725	JFK
2013	1	1	LGA	2013-01-01	-6	461	LGA
2013	1	1	EWR	2013-01-01	-4	1696	EWR

# filter by airport
dEWR <- dd %>% filter(origin == "EWR")
dLGA <- dd %>% filter(origin == "LGA")
dJFK <- dd %>% filter(origin == "JFK")

# minimum delays --> departure ahead of time

# group by date
dEWR <- dEWR %>% group_by(date)
dLGA <- dLGA %>% group_by(date)
dJFK <- dJFK %>% group_by(date)

dEWRmin <- filter(dEWR, depDelay <= min(depDelay))
dLGAmin <- filter(dLGA, depDelay <= min(depDelay))
dJFKmin <- filter(dJFK, depDelay <= min(depDelay))

# maximum delays

dEWRmax <- filter(dEWR, depDelay >= max(depDelay))
dLGAmax <- filter(dLGA, depDelay >= max(depDelay))
dJFKmax <- filter(dJFK, depDelay >= max(depDelay))

# average 

dEWRmean <- dEWR %>% summarise( meanDelay = mean(depDelay))
dLGAmean <- dLGA %>% summarise( meanDelay = mean(depDelay))
dJFKmean <- dJFK %>% summarise( meanDelay = mean(depDelay))

# plot for each airport

gEWR <- ggplot() + 
        geom_line(data = dEWRmean, aes(x = date, y = meanDelay, color = "Mean"))

gEWR <- gEWR + 
        geom_line(data = dEWRmax, aes(x = date, y = depDelay, color = "Maximum"))

gEWR <- gEWR +  
        geom_line(data = dEWRmin, aes(x = date, y = depDelay, color = "Minimum"))

gEWR <- gEWR + 
        theme_bw() + 
        theme( legend.position = "right",
        plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Days", 
              y = "Delay (min/day)", 
              title = "Departure Delay per Day for Flights from EWR Airport in NYC (2013)" )

gEWR <- gEWR +
        scale_x_date(date_breaks = "2 month", date_labels = "%b-%d") +
        scale_y_continuous( n.breaks = 8 ) +
        labs(color='Delay')

# plot LGA

gLGA <- ggplot() + 
        geom_line(data = dLGAmean, aes(x = date, y = meanDelay, color = "Mean"))

gLGA <- gLGA + 
        geom_line(data = dLGAmax, aes(x = date, y = depDelay, color = "Maximum"))

gLGA <- gLGA +  
        geom_line(data = dLGAmin, aes(x = date, y = depDelay, color = "Minimum"))

gLGA <- gLGA + 
        theme_bw() + 
        theme( legend.position = "right",
        plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Days", 
              y = "Delay (min/day)", 
              title = "Departure Delay per Day for Flights from LGA Airport in NYC (2013)" )

gLGA <- gLGA +
        scale_x_date(date_breaks = "2 month", date_labels = "%b-%d") +
        scale_y_continuous( n.breaks = 8 ) +
        labs(color='Delay')

# plot JFK

gJFK <- ggplot() + 
        geom_line(data = dJFKmean, aes(x = date, y = meanDelay, color = "Mean"))

gJFK <- gJFK + 
        geom_line(data = dJFKmax, aes(x = date, y = depDelay, color = "Maximum"))

gJFK <- gJFK +  
        geom_line(data = dJFKmin, aes(x = date, y = depDelay, color = "Minimum"))

gJFK <- gJFK + 
        theme_bw() + 
        theme( legend.position = "right",
        plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Days", 
              y = "Delay (min/day)", 
              title = "Departure Delay per Day for Flights from JFK Airport in NYC (2013)" )

gJFK <- gJFK +
        scale_x_date(date_breaks = "2 month", date_labels = "%b-%d") +
        scale_y_continuous( n.breaks = 8 ) +
        labs(color='Delay')

gEWR
gLGA
gJFK

3) Average speed of each plane

s <- flights %>% transmute( date = paste(year, month, day, sep = "-"),
                            duration = air_time / 60, # hours
                            distance = distance * 1.609,
                            plane = tailnum) # km

s <- s %>% transmute( date = as.Date(date),
                      avSpeed = distance / duration,
                      plane = plane)
s <- s %>% filter(!is.na(avSpeed))

s <- s %>% group_by(date)
s <- s %>% summarise(mean = mean(avSpeed))
head(s)

A tibble: 6 × 2

date	mean
<date>	<dbl>
2013-01-01	575.0771
2013-01-02	587.5402
2013-01-03	594.9819
2013-01-04	616.8570
2013-01-05	604.2606
2013-01-06	593.3873

# plot

ps <- ggplot(s, aes( x = date, y = mean) ) + 
      geom_line( linewidth = 0.8, color = "#58508d") 

ps <- ps +
        theme_bw() + 
        theme( plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        labs( x = "Days", 
              y = "Average Plane Speed (km/h)", 
              title = "Average Plane Speed per Day for Flights from NYC Airports in 2013" )

ps <- ps + 
      scale_x_date(date_breaks = "2 month", date_labels = "%b-%d") +
      scale_y_continuous( n.breaks = 8 )

ps

4.1) Airline companies offering the largest two numbers of flights per day and per week

w <- flights %>% transmute( date = paste(year, month, day, sep = "-"),
                            flight = flight,
                            Airline = carrier)
w <- w %>% transmute( date = as.Date(date),
                      week = isoweek(ymd(date)),
                      flight = flight,
                      Airline = Airline)

# per day

wd <- w %>% group_by(date, Airline)
wd <- wd %>% transmute( date = date,
                        totDay = length(flight),
                        Airline = Airline)

# finding the largest two numbers

wd <- filter(wd, totDay >= max(totDay))
wd <- wd %>% slice(1:1)
wd <- wd %>% ungroup()
wd <- wd %>% arrange(desc(totDay)) %>% group_by(date) %>% slice(1:2)

head(wd)

A grouped_df: 6 × 3

date	totDay	Airline
<date>	<int>	<chr>
2013-01-01	165	UA
2013-01-01	163	B6
2013-01-02	170	UA
2013-01-02	162	B6
2013-01-03	162	B6
2013-01-03	159	UA

# plot

m <- ggplot(wd, aes( x = date, y = totDay, color = Airline) ) + 
      geom_point( size = 2 ) 

m <- m +
        theme_bw() + 
        theme( plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Days", 
              y = "Number of Flights", 
              title = "Two Largest Numbers of Flights per Day - Flights From NYC Airports, 2013" )

m <- m + 
      scale_x_date(date_breaks = "1 month", date_labels = "%m/%d") +
      scale_y_continuous( n.breaks = 8 )

m

# week

ww <- w %>% group_by(week, Airline)
ww <- ww %>% transmute( week = week,
                        totWeek = length(flight),
                        Airline = Airline)

ww <- filter(ww, totWeek >= max(totWeek))
ww <- ww %>% slice(1:1)
ww <- ww %>% ungroup()
ww <- ww %>% arrange(desc(totWeek)) %>% group_by(week) %>% slice(1:2)

head(ww)

A grouped_df: 6 × 3

week	totWeek	Airline
<dbl>	<int>	<chr>
1	1284	B6
1	1230	UA
2	1035	UA
2	994	B6
3	1032	UA
3	970	B6

# plot

mm <- ggplot(ww, aes( x = week, y = totWeek, color = Airline) ) + 
      geom_point( size = 3)

mm <- mm +
        theme_bw() + 
        theme( plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Week Number", 
              y = "Number of Flights", 
              title = "Two Largest Numbers of Flights per Week - Flights From NYC Airports, 2013" )

mm <- mm + 
      scale_x_continuous( n.breaks = 10 ) +
      scale_y_continuous( n.breaks = 4 )

mm

4.2) Airline company offering the smallest number of flight per month

t <- flights %>% transmute( month = month.abb[month],
                            flight = flight,
                            Airline = carrier)
t <- t %>% group_by(month, Airline)
t <- t %>% transmute( month = month,
                      totMonth = length(flight),
                      Airline = Airline)
t <- filter(t, totMonth <= min(totMonth))
t <- t %>% slice(1:1)
t <- t %>% ungroup()
t <- t %>% arrange(desc(totMonth)) %>% group_by(month) %>% slice(1:1)
head(t)

A grouped_df: 6 × 3

month	totMonth	Airline
<chr>	<int>	<chr>
Apr	5047	UA
Aug	5124	UA
Dec	4931	UA
Feb	4346	UA
Jan	4637	UA
Jul	5066	UA

# plot

z <- ggplot(t, aes( x = month, y = totMonth, color = Airline) ) + 
      geom_point( size = 4 ) 

z <- z +
        theme_bw() + 
        theme( plot.title = element_text(size = 22, hjust = 0.5),
        axis.title = element_text(size = 18),
        axis.text  = element_text(size = 16),
        legend.title = element_text(size = 20, hjust = 0.5),
        legend.text  = element_text(size = 18),
        panel.grid.major.y = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.major.x = element_line( color = alpha("black",0.2),
                                           linewidth = 0.3,
                                           linetype = 2 ),
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        legend.background = element_rect( fill = "white",
                                          linewidth = 0.5, 
                                          linetype = "solid", 
                                          colour = "grey" ) ) +
        scale_color_manual(values = color_palette) +
        labs( x = "Month", 
              y = "Number of Flights", 
              title = "Smallest Numbers of Flights per Month - Flights From NYC Airports, 2013" )

z <- z + 
      scale_y_continuous( n.breaks = 7 )

z

4.3) Airline company offering the longest distance flight per month

q <- flights %>% transmute( month = month.abb[month],
                            Airline = carrier,
                            distance = distance * 1.609) # km
q <- q %>% group_by(month)
q <- filter(q, distance >= max(distance))
q <- q %>% slice(1:1)
q

A grouped_df: 12 × 3

month	Airline	distance
<chr>	<chr>	<dbl>
Apr	HA	8017.647
Aug	HA	8017.647
Dec	HA	8017.647
Feb	HA	8017.647
Jan	HA	8017.647
Jul	HA	8017.647
Jun	HA	8017.647
Mar	HA	8017.647
May	HA	8017.647
Nov	HA	8017.647
Oct	HA	8017.647
Sep	HA	8017.647